Why do women and men compare two-digit numbers differently?
Women were more thrown off than men when the tens and units digits of two numbers pointed in opposite directions, and this difference was explained by how people navigated, not by their hormone levels.
Source
Sex Differences in Number Magnitude Processing Strategies Are Mediated by Spatial Navigation Strategies: Evidence From the Unit-Decade Compatibility Effect
Study at a glance
- Design
- Cross-sectional — Men and naturally cycling women tested once (women in the mid-luteal phase) on a two-digit number comparison task that manipulated unit-decade compatibility and vertical spacing, plus a 3-D virtual navigation task and salivary hormone measures; sex differences, hormone links and a mediation by navigation strategy were tested with mixed models.
- N
- N=83 · After exclusions, 42 men and 41 women were analysed; hormone values were missing for four participants.
- Population
- Healthy, right-handed, IQ- and age-matched young adults aged 18 to 35 at the University of Salzburg; women not using hormonal contraception
- Outcome
- Unit-decade compatibility effect (incompatible minus compatible) in reaction time and error rate; landmark vs. distance strategy effect in navigation time; salivary estradiol, progesterone and testosterone
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
Men and women matched for age and IQ decided which of two vertically stacked two-digit numbers was larger, with some pairs 'compatible' (the larger number also had the larger units digit) and some 'incompatible', and with the numbers placed either close together or far apart. The same people navigated virtual environments using directions based either on landmarks or on distances, and gave saliva samples for hormone measurement. Women were all tested in the high-progesterone luteal phase.
What they found
Women made more errors than men overall, mainly on incompatible pairs, giving them a larger compatibility effect in errors (5.49% vs. 3.10%), but there was no sex difference in reaction times. Closely spaced numbers produced a larger compatibility effect in reaction time for everyone, and spacing did not interact with sex. Individual hormone levels did not predict the error compatibility effect, and the progesterone-to-testosterone ratio showed only a small link that did not explain the sex difference. People who did relatively better with landmark directions had larger compatibility effects, and navigation strategy statistically mediated the sex difference.
The limits
What it doesn't show
Because the design is correlational, the mediation analysis cannot show that navigation style causes number-processing style; both may reflect a shared tendency towards local or global processing. The sex difference appeared only in errors, not in reaction times, unlike most earlier studies, and the authors acknowledge limited power to detect a sex-by-spacing interaction. Women were tested only in the luteal phase, so cycle effects cannot be assessed, and the navigation results they rely on were reported in an unpublished paper. The text is also inconsistent about whether the hormone-ratio effect was in reaction times or errors.
Key terms
- Unit-decade compatibility effect
- Slower or less accurate comparisons when the number with larger tens has smaller units (e.g., 61 vs. 29) than when both digits agree (e.g., 69 vs. 21).
- Holistic vs. decomposed number processing
- Comparing two-digit numbers as whole magnitudes versus comparing tens and units separately; a bigger compatibility effect suggests more decomposed processing.
- Mental number line
- The idea that numbers are represented spatially along a line, with smaller numbers more spread out than larger ones.
- Global-local processing
- The tendency to attend to the overall shape of a pattern versus its component details.
- Mediation analysis
- A statistical test of whether the link between two variables runs partly through a third variable; it does not by itself establish causation.
Flashcards
0 of 9 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
Which pair is an incompatible item?
Common questions
Does a larger compatibility effect mean women are worse at maths?
No. It indicates a tendency to process tens and units separately; overall speed did not differ, and the difference was a few percentage points in errors on one item type.
Why test women only in the luteal phase?
Progesterone peaks then, and progesterone had been linked to more detail-focused processing, so this was the phase where sex differences were expected to be largest.
If navigation strategy 'mediates' the sex difference, does learning a different navigation style change number processing?
Not necessarily. The data are correlational, so both may simply reflect a common underlying preference for local detail rather than one causing the other.
More on Numerical cognition